Merge pull request #5536 from janezd/distribution-deepcopy

[FIX] Fix deepcopy and pickle for classes derived from `np.ndarray`

Orange/statistics/contingency.py

@@ -179,9 +179,20 @@ def __reduce__(self):
         return (
             _create_discrete,
             (Discrete, np.copy(self), self.col_variable, self.row_variable,
-             self.col_unknowns, self.row_unknowns)
+             self.col_unknowns, self.row_unknowns, self.unknowns)
         )
 
+    def __array_finalize__(self, obj):
+        # defined in __new__, pylint: disable=attribute-defined-outside-init
+        """See http://docs.scipy.org/doc/numpy/user/basics.subclassing.html"""
+        if obj is None:
+            return
+        self.col_variable = getattr(obj, 'col_variable', None)
+        self.row_variable = getattr(obj, 'row_variable', None)
+        self.col_unknowns = getattr(obj, 'col_unknowns', None)
+        self.row_unknowns = getattr(obj, 'row_unknowns', None)
+        self.unknowns = getattr(obj, 'unknowns', None)
+
 
 class Continuous:
     def __init__(self, dat, col_variable=None, row_variable=None,

Orange/statistics/distribution.py

@@ -30,6 +30,26 @@ def _get_variable(dat, variable, expected_type=None, expected_name=""):
 
 
 class Distribution(np.ndarray):
+    def __array_finalize__(self, obj):
+        # defined in derived classes,
+        # pylint: disable=attribute-defined-outside-init
+        """See http://docs.scipy.org/doc/numpy/user/basics.subclassing.html"""
+        if obj is None:
+            return
+        self.variable = getattr(obj, 'variable', None)
+        self.unknowns = getattr(obj, 'unknowns', 0)
+
+    def __reduce__(self):
+        state = super().__reduce__()
+        newstate = state[2] + (self.variable, self.unknowns)
+        return state[0], state[1], newstate
+
+    def __setstate__(self, state):
+        # defined in derived classes,
+        # pylint: disable=attribute-defined-outside-init
+        super().__setstate__(state[:-2])
+        self.variable, self.unknowns = state[-2:]
+
     def __eq__(self, other):
         return (
             np.array_equal(self, other) and

Orange/tests/test_contingency.py

@@ -1,6 +1,6 @@
 # Test methods with long descriptive names can omit docstrings
 # pylint: disable=missing-docstring
-
+import copy
 import unittest
 from unittest.mock import Mock
 
@@ -71,6 +71,13 @@ def test_discrete_missing(self):
         np.testing.assert_almost_equal(cont.row_unknowns, [0, 0])
         self.assertEqual(1, cont.unknowns)
 
+    def test_deepcopy(self):
+        cont = contingency.Discrete(self.zoo, 0)
+        dc = copy.deepcopy(cont)
+        self.assertEqual(dc, cont)
+        self.assertEqual(dc.col_variable, cont.col_variable)
+        self.assertEqual(dc.row_variable, cont.row_variable)
+
     def test_array_with_unknowns(self):
         d = data.Table("zoo")
         d.Y[2] = float("nan")

Orange/tests/test_distribution.py

@@ -1,7 +1,8 @@
 # Test methods with long descriptive names can omit docstrings
 # Test internal methods
 # pylint: disable=missing-docstring, protected-access
-
+import copy
+import pickle
 import unittest
 from unittest.mock import Mock
 import warnings
@@ -110,6 +111,32 @@ def test_fallback_with_weights_and_nan(self):
             np.asarray(fallback), np.asarray(default))
         np.testing.assert_almost_equal(fallback.unknowns, default.unknowns)
 
+    def test_pickle(self):
+        d = data.Table("zoo")
+        d1 = distribution.Discrete(d, 0)
+        dc = pickle.loads(pickle.dumps(d1))
+        # This always worked because `other` wasn't required to have `unknowns`
+        self.assertEqual(d1, dc)
+        # This failed before implementing `__reduce__`
+        self.assertEqual(dc, d1)
+        self.assertEqual(hash(d1), hash(dc))
+        # Test that `dc` has the required attributes
+        self.assertEqual(dc.variable, d1.variable)
+        self.assertEqual(dc.unknowns, d1.unknowns)
+
+    def test_deepcopy(self):
+        d = data.Table("zoo")
+        d1 = distribution.Discrete(d, 0)
+        dc = copy.deepcopy(d1)
+        # This always worked because `other` wasn't required to have `unknowns`
+        self.assertEqual(d1, dc)
+        # This failed before implementing `__deepcopy__`
+        self.assertEqual(dc, d1)
+        self.assertEqual(hash(d1), hash(dc))
+        # Test that `dc` has the required attributes
+        self.assertEqual(dc.variable, d1.variable)
+        self.assertEqual(dc.unknowns, d1.unknowns)
+
     def test_equality(self):
         d = data.Table("zoo")
         d1 = distribution.Discrete(d, 0)
@@ -285,6 +312,30 @@ def test_construction(self):
         self.assertEqual(disc2.unknowns, 0)
         assert_dist_equal(disc2, dd)
 
+    def test_pickle(self):
+        d1 = distribution.Continuous(self.iris, 0)
+        dc = pickle.loads(pickle.dumps(d1))
+        # This always worked because `other` wasn't required to have `unknowns`
+        self.assertEqual(d1, dc)
+        # This failed before implementing `__reduce__`
+        self.assertEqual(dc, d1)
+        self.assertEqual(hash(d1), hash(dc))
+        # Test that `dc` has the required attributes
+        self.assertEqual(dc.variable, d1.variable)
+        self.assertEqual(dc.unknowns, d1.unknowns)
+
+    def test_deepcopy(self):
+        d1 = distribution.Continuous(self.iris, 0)
+        dc = copy.deepcopy(d1)
+        # This always worked because `other` wasn't required to have `unknowns`
+        self.assertEqual(d1, dc)
+        # This failed before implementing `__deepcopy__`
+        self.assertEqual(dc, d1)
+        self.assertEqual(hash(d1), hash(dc))
+        # Test that `dc` has the required attributes
+        self.assertEqual(dc.variable, d1.variable)
+        self.assertEqual(dc.unknowns, d1.unknowns)
+
     def test_hash(self):
         d = self.iris
         petal_length = d.columns.petal_length